Improve Cypress teardown

[martinjagodic]

Improve logging for various Cypress tasks and streamline API request handling. Adjust timeout settings to enhance test reliability and performance. This update addresses issues with task timeouts and logging verbosity, ensuring a more efficient testing process.

[yanthomasdev]

Tests are hanging for no apparent reason.

I think it might be related to GitHub itself, there's some ungoing issues with Actions, I'll see if there's any issues happening locally

[yanthomasdev]

Looks good to me, I have a small nit but not blocking.

[martinjagodic]

@yanthomasdev I think this was not a GA problem, because the pattern is always the same - the last of 4 parallel machines hangs at teardown. This has been happening for a while now, but not on every run.

If you have any other proposal on how to run tests, I would be glad to hear it

[Copilot]

Pull request overview

This PR aims to reduce Cypress CI hangs (especially around git-gateway setup/teardown) by adding diagnostics, tightening timeouts, and adjusting Cypress time/clock behavior.

Changes:

• Added CI logging and reduced the execa() Cypress runner timeout.
• Added request timeouts + retry adjustments for Netlify API calls used by the git-gateway Cypress plugin.
• Skipped cy.clock() for git-gateway tests to avoid hangs during page load.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
cypress/utils/steps.js	Refactors several Cypress command sequences; adds an eslint disable and changes some chained commands to repeated queries.
cypress/run.mjs	Adds fork/machine debug logging and changes Cypress runner timeout behavior.
cypress/plugins/gitGateway.js	Consolidates Netlify API calls into a timed fetch helper and adjusts deploy polling + teardown delete behavior.
cypress/e2e/common/spec_utils.js	Changes hook exports to functions, lowers task timeout, and conditionally skips cy.clock() for git-gateway.
cypress.config.ts	Removes/adjusts comments around retries/plugins configuration.
.github/workflows/nodejs.yml	Adds a clarifying comment for fork-only env usage.

Comments suppressed due to low confidence (1)

cypress/run.mjs:25

• Manual sharding can break when specs.length < machineCount: Math.floor(specs.length / machineCount) becomes 0, causing most machines to run 0 specs and the last machine to run everything. Consider using Math.ceil (and clamping start/end) or guarding specsPerMachine to be at least 1 to keep shards balanced.

  if (isFork && machineIndex && machineCount) {
    const specsPerMachine = Math.floor(specs.length / machineCount);
    const start = (machineIndex - 1) * specsPerMachine;
    const machineSpecs =
      machineIndex === machineCount
        ? specs.slice(start)
        : specs.slice(start, start + specsPerMachine);

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[yanthomasdev]

I have one small comment but I think it's close to getting merged

[yanthomasdev]

Don't we want to fetch this with fetchWithTimeout too?